Prediction of Terrorist Activities by Using Unsupervised Learning Techniques

Terrorism now considered as a major threat to the world population. Terrorism is increasing day by day by different means. From the last decade terrorism rate is increasing exponentially. But there is no efficient way for prediction of terrorism activities. Our paper focuses on prediction of different terrorist attacks by using data mining techniques. In this paper we proposed prediction of attacks by using unsupervised clustering technique. We proposed a framework in which we do sentiments analysis of our data and then by using a combination of density based clustering and distance based clustering we assign classes to our data. Class labels help us to predict terrorism attacks. By research we come to know that combination of these two clustering techniques give accurate results. This proposed framework gives high level of accuracy and it is useful in prediction of attacks types. It gives us a way to deal with terrorism attacks in advance and makes our society peaceful.